1. Field of the Invention
The present invention relates to an optical-switch testing apparatus having a microelectromechanical systems (MEMS) mirror array, and more particularly, to an optical-switch testing apparatus and an optical-switch testing method to obtain a deflection control amount with efficiency and high precision, an optical-signal switching apparatus equipped with the optical switch to stabilize light output, and a control method for the optical-signal switching apparatus.
2. Description of the Related Art
Recently, along with the rapid spread of internet, traffic volume is remarkably increasing. One of the methods to cope with the increasing traffic and to configure a large capacity optical communication nets is a wavelength division multiplexing (WDM) system. A backbone optical network based on the WDM system is an optical cross-connect (OXC) system. The OXC system has a structure in which plural optical-signal switching apparatuses are connected mutually on light transmission paths consisting of optical fibers.
WDM signals are input to the optical-signal switching apparatus via optical fibers, and the wavelength of optical signals in a same light transmission path is transmitted in multiplexed manners. Optical switches are arranged inside the optical-signal switching apparatus, and light paths of the optical signals are changed to other light paths in unit of wavelength by the optical switches, and may be output to other light transmission paths.
According to the OXC system employing optical-signal switching apparatuses, when an optical fiber configuring a certain light transmission path becomes faulty, light paths are switched by the internal optical switch instantaneously, and thereby optical signals are diverted to an optical fiber configuring other light transmission path or an optical fiber in other route. Accordingly, even when a trouble occurs with a light transmission path, a recovery can be made at high speed, and light paths can be switched in unit of wavelength.
FIG. 13 is a perspective view of an optical switch. An optical switch 100 of 3-dimensional type MEMS includes two MEMS mirror arrays 101 and 102 manufactured by application of Bulk micromachine technology, and two collimator arrays 103 and 104 that make light go into and go out from these two MEMS mirror arrays 101 and 102. The MEMS mirror arrays 101 and 102 are manufactured by etching raw material boards and forming their supporting bodies and mirror forming portions into desired shapes, and by forming mirror surfaces and electrodes into thin films. The use of the MEMS mirror arrays 101 and 102 enables to structure an optical switch that changes light paths spatially in 3-dimensional manners.
The MEMS mirror arrays 101 and 102 includes plural tilt mirrors 105 and 106 in matrix shape, respectively. The angle of each of the tilt mirrors 105 and 106 may be changed independently in two axis directions (X and Y), and by changing the incident angle of incoming light A, the optical paths of the light A may be changed to arbitrary angles. In the collimator arrays 103 and 104, a number of input and output ports corresponding to the plural tilt mirrors 105 and 106 of the MEMS mirror arrays 101 and 102 are formed in matrix shape, respectively.
The technology concerning such a 3-dimensional type MEMS optical switch is disclosed, for example, in “High Speed Switching 3-Dimensional Type MEMS Optical Switch”, technical digest of Communication Society Conference, p. 447, IEICE, 2002. In addition, the technology concerning MEMS mirrors that have sinking comb type electrodes enabling to change angles of respective mirrors in two axis directions is disclosed, for example, in Japanese Patent Application Laid-Open Publication No. 2002-328316.
The optical switch employing such MEMS mirror arrays 101 and 102 is superior to other switches in viewpoints of compact size, wavelength independency, polarized wave independency and other features, therefore attracts much attention of those skilled in the art. In addition, an optical-signal switching apparatus employing such an optical switch 100 of 3-dimensional type MEMS as mentioned above realizes reduction of light loss, large capacity, and multiple channels.
In the optical switch 100, to the collimator arrays 103 and 104, end portions of optical fibers (not shown) of plural ports (103-1 to 103-N, 104-1 to 104-N) are arranged, respectively. In the example shown in the figure, the collimator array 103 at the input side and the collimator array 104 at the output side are arranged in parallel so that the respective light input directions should face forward in the figure, while the respective light output directions should face backward in the figure.
The collimator arrays 103 and 104 have plural ports (103-1 to 103-N, 104-1 to 104-N) in matrix shape in vertical and horizontal directions, respectively. To each of the plural ports, an end portion of an optical fiber (not shown) is fixedly arranged for inputting and outputting optical signals. The backward surface of each port of the collimator arrays 103 and 104 is end surface processed for making light go out from an optical fiber.
At the back of the collimator arrays 103 and 104, the MEMS mirror arrays 101 and 102 are arranged in correspondence to the arrangement interval between the collimator arrays 103 and 104. The MEMS mirror arrays 101 and 102 are arranged so as to be respectively inclined by 45 degrees to the direction of the light path A between the collimator arrays 103 and 104. In addition, the collimator arrays 103 and 104 are arranged so as to be at right angles to each other. The light input to each port of the collimator array 103 at the input side is made to go out as a light path A, and by the MEMS mirror array 101 at the input side, the light path A is reflected toward the MEMS mirror array 102 at the output side. Thereafter, by the MEMS mirror array 102, the light path A is reflected toward the collimator array 104, and may be output through each port of the collimator array 104.
The input mirror 105 and the output mirror 106 include a deflecting unit (not shown) having the sinking comb type electrode. By supplying deflection control amounts (driving voltage) corresponding to angle changes to this deflecting unit, the angles of the input mirror 105 and the output mirror 106 are continuously changed in correspondence to the values of the driving voltage.
At the time of a through pass, the light A that is output from the port 3 (103-3) of the collimator array 103 is reflected by the input mirror 105-3 of the MEMS mirror array 101, then reflected by the output mirror 106-3 of the MEMS mirror array 102, and is made to go into the port 3 (104-3) of the collimator array 104. At this moment, the surfaces of the input mirror 105-3 of the MEMS mirror array 101 and the output mirror 106-3 of the MEMS mirror array 102 are in a parallel status with the surfaces of the main bodies of the MEMS mirror arrays 101 and 102. In this status, angle change control to the input mirror 105-3 and the output mirror 106-3 is not carried out.
At the time of a cross pass, angle change control to the input mirror 105-3 of the MEMS mirror array 101 and the output mirror 106-3 of the MEMS mirror array 102 is carried out, thereby, the reflection direction of the incident light of the light A to the port 3 (103-3) of the collimator array 103 is deflected, and the light is made to go into an arbitrary port (one of 104-1 to 104-N) of the collimator array 104. In this manner, the optical switch 100 enables to change light input from plural ports to an arbitrary port and output light thereto. An optical-signal switching apparatus to be described later herein is the device that switches optical signals of plural systems on a light transmission path by this port change.
While, in the optical switch 100, in both a through pass and a cross pass, the light path (the optical axis of the light A) may go with displacement into an optical fiber at the output side connected to the collimator array 104 at the output side. This displacement of a light path will occur from factors including the structural characteristics of the MEMS mirror arrays 101 and 102, angle displacement of a mirror that actually works to the controlled variables at the moment of angle change, displacement of component arrangement at assembly, and so forth. Further, this light path displacement becomes a factor to increase light loss of an optical-signal switching apparatus equipped with the optical switch 100. Control technologies to decrease light loss in such an optical switch are disclosed, for example, in Japanese Patent Application Laid-Open Publication No. H9-508218 and Japanese Patent Application Laid-Open Publication No. 2002-236264.
FIG. 14 is a flowchart of a test procedure for obtaining a deflection control amount of an optical switch according to a conventional technology. The test is carried out on the assumption that there is no displacement of component arrangement at assembly of the optical switch 100. By the way, for the sake of convenience, the following explanation is made with the collimator array 103 side as the light input side, with the collimator array 104 as the light output side, and with the tilt mirror 105 of the MEMS mirror array 101 as an input mirror, and with the tilt mirror 106 of the MEMS mirror array 102 as an output mirror.
In the first place, in the setting of the input and output ports, an input port of the collimator array 103 and an output port of the collimator array 104 are set (step S101). Then, a deflection angle with which light should be output from the output port set at the collimator array 104 side, when light is input from the input port set at the collimator array 103 side, is calculated (step S102). This deflection angle is calculated on the basis of the theoretical values necessary for angle changes of the input mirror 105 of the MEMS mirror array 101 and the output mirror 106 of the MEMS mirror array 102.
In the next place, by use of the calculated deflection angle, angle change control is carried out to the respective deflecting units (not shown) of the input mirror 105 of the MEMS mirror array 101 and the output mirror 106 of the MEMS mirror array 102. In concrete, the deflecting units are driven (step S103), and thereby, the angles of the two axes (X and Y) of the input mirror 105 and the output mirror 106 are changed in continuous manners. At the same time, the light receiving level of the light A that is output from the output port of the collimator array 104 is detected by a light detector (not shown) (step S104), and the optimal point of light loss is obtained (step S105).
Until the optimal point is obtained (step S105: No), the driving of the deflecting units by the step S103, and the detection of the light receiving level by the step S104 are continued by feedback control. When the optimal point is obtained (step S105: Yes), the angle position corresponding to this optimal point is taken as a deflection control amount. The optimal point is the angle position at the moment when the light loss between the previously set input and output ports becomes minimum, namely, when the light receiving level is detected maximum.
Through the process mentioned above, the deflection control amounts at the input port and the output port set in the step S101, that is, the input port of the collimator array 103, and the output port of the collimator array 104, are obtained. Thereafter, a new combination of an input port and an output port is set, and the process mentioned above is carried out thereon (step S106: No). When the optimal points are obtained for all the ports of the collimator arrays 103 and 104 in matrix shapes (step S106: Yes), the deflection control amounts for all the combinations of input and output ports by the through pass and cross pass are obtained.
However, in the conventional art, in the instance of displacement of component arrangement at assembly of the optical switch 100, the deflection control amount at which the light receiving level of output becomes maximum will shift from the theoretical value. As a consequence, the process to obtain the deflection control amount at which the light receiving level becomes maximum, in other words, the number of times of feedback control in the steps S103-S105 will increase, and test time will increase accordingly, which has been a problem with the conventional art.
Further, in the test process mentioned above, the deflection control amount is calculated by use of the theoretical values of the MEMS mirror arrays 101 and 102. Accordingly, when there is conspicuous fluctuation peculiar to each of the input mirror 105 of the MEMS mirror array 101 and the output mirror 106 of the MEMS mirror array 102, the displacement in the calculated value of the deflection control amount will become large, and test time will increase accordingly, which has been another problem with the conventional art.
When the number of the input ports of an optical switch is N, and the number of the output ports thereof is N, the number of changeable combinations by a through pass and a cross pass is N2 ways. In the optical switch 100 shown in FIG. 13, because displacement in component arrangement occurs at assembly, when the number of the input port is defined as i, and the number of the output port is defined as j, even in a test of a through pass in which it stands that i=j, deflection control must be executed. In addition, with regard to this displacement of component arrangement that occurs at assembly, even in a test of a cross pass in which it stands that i≠j, too, deflection control must be executed. Therefore, according to the conventional test process, because the time for testing a certain pass is around five minutes, in order to test the deflection control amount of an optical switch of N×N, test time of 5×N2 minutes is required as a whole.
While, the optical switch testing process mentioned above is carried out by use of an optical-switch testing apparatus. An optical-switch testing apparatus includes a light source that outputs optical signals to respective input ports of an optical switch, a light detector that detects the light receiving levels of optical signals obtained from respective output ports of the optical switch, and a control circuit that executes the testing process mentioned above. The control circuit memorizes the deflection control amounts obtained by the testing process mentioned above. Accordingly, the deflection control amounts obtained as the results of testing the optical switch are stored into a memory as control parameters peculiar to each optical switch.
The optical-signal switching apparatus into which an optical switch is assembled reads parameters stored in this memory, and thereby operates actual optical signal change actions. For example, when the optical-signal switching apparatus receives a change instruction, it reads the deflection control amounts corresponding to the set input and output ports from the memory and thereby drives the deflecting units.
FIG. 15 is a table of memory contents to store the deflection control amount. The contents of parameters that are stored in a memory 110 are shown in the table. The number of combinations of an input port and an output port is N2 ways as mentioned previously. Further, because the deflection control amounts of the X axis and the Y axis for the input mirror 105 of the MEMS mirror array 101 and the output mirror 106 of the MEMS mirror array 102 are required for one pass, therefore, four pieces of deflection control amounts are necessary. As a consequence, in the instance of an optical switch having the number of input ports N and the number of output ports N, it is required to store 4N2 pieces of deflection control amounts. Therefore, in the conventional art, 4N2 pieces of deflection control amounts had to be stored into the memory 110, the memory capacity increased inevitably. Especially, in the instance when the scale of an optical switch becomes large in future owing to an increased number of input and output ports, because the necessary memory capacity increases by the square of the number of ports, the memory capacity will increase exponentially.